Volatile Origin and Cycles: Nitrogen and Methane, S. K. Atreya, R. D. Lorenz, and J. H. Waite, in Titan from Cassini-Huygens (R.H. Brown et al.

Transcription

1 Volatile Origin and Cycles: Nitrogen and Methane, S. K. Atreya, R. D. Lorenz, and J. H. Waite, in Titan from Cassini-Huygens (R.H. Brown et al., eds.), Chapter 7, pp , 2009, DOI / _7, Springer Dordrecht, Heidelberg-London-New York.

2 Chapter 7 Volatile Origin and Cycles: Nitrogen and Methane Sushil K. Atreya, Ralph D. Lorenz, and J. Hunter Waite Abstract The story of Titan s two most abundant volatile constituents, nitrogen and methane, is intertwined. The focus of this paper is the origin and evolution of Titan s nitrogen atmosphere and the cycle of methane from its production to destruction to replenishment. Relevant observational results from Cassini Huygens, Voyager and the Earth as well as various hypotheses and models are reviewed. The origin of nitrogen by direct capture, and from dissociation of primordial nitrogen-bearing molecules, especially ammonia, by impact, photolysis, thermal and other processes is evaluated. Similarly, the origin of methane from Saturn s sub-nebula or by water-rock reactions in Titan s interior is reviewed. The role of methane in regulating Titan s climate is noted, and similarities and differences between the methane cycle in Titan s troposphere and the hydrological cycle on Earth are discussed. The fate of methane in the stratosphere and the upper atmosphere/ionosphere is examined in order to evaluate the possibility and extent of an ocean of ethane, requirement of methane replenishment, and the role of product aerosols in maintaining Titan s nitrogen atmosphere. 7.1 Historical Perspective: From Christiaan Huygens to Cassini Huygens Titan is unique in the solar system. Whereas satellites in general are not known for atmospheres, Titan is not only endowed with an atmosphere, it has a massive atmosphere with surface pressure exceeding that on Earth by 50%. Moreover, its atmosphere is largely made up of nitrogen (N 2 ). Its second most abundant constituent, methane (CH 4 ), S.K. Atreya ( ) University of Michigan, Ann Arbor, MI, , USA atreya@umich.edu R.D. Lorenz Johns Hopkins University, Applied Physics Laboratory, Laurel, MD, 20723, USA J. Hunter Waite South West Research Institute, San Antonio, TX, 78228, USA displays a meteorological cycle similar to the hydrological cycle on Earth, complete with surface reservoirs, evaporation, clouds, rain, and dendritic channels presumably carved by the rain of methane. In Titan s stratosphere and the ionosphere, the two volatiles, nitrogen and methane, together create a complex web of chemical reactions. The resulting trace species condense or otherwise produce multiple layers of aerosols, filling Titan s atmosphere with photochemical smog and haze. The aerosols play a crucial role on Titan, however. They provide stratospheric warming, which is critical in preventing nitrogen from condensing. The above modern-day view of Titan had its origins in 1655 when Christiaan Huygens discovered Titan a large satellite, in geosynchronous orbit at Saturn radii from the planet Saturn. First indication of an atmosphere around Titan began to emerge in José Comas Solá recorded limb darkening, which he attributed to the existence of a strongly absorbing atmosphere around Titan. This was a remarkable finding for the time, considering Titan s angular size is only 0.1 arc sec, which is at the limit of the seeing capability of best ground-based telescopes. Yet Comas Solá was gifted with exceptional acuity, and his sketches (e.g. Lorenz 1997) of the Galilean satellites are broadly accurate, and he knew that Mars had no canals. The first quantitative spectroscopic observations of Titan began in earnest in 1944 with Gerard Kuiper s detection of methane at 6190 Å in the visible and in the near-infrared. Kuiper (1944) derived 0.2 km-am for the methane abundance. Trafton (1972) detected methane in the 3 n 3 band of CH 4 at 1.1 m m, and arrived at a much greater abundance of 1.6 km-am, or 16 mb, considering self-broadening in a pure CH 4 atmosphere. An even greater surface pressure of 200 mb was suggested by Lutz et al. (1976) from their analysis of the weak methane bands in the visible where pressure broadening is not a complicating factor. However, they concluded that methane is a minor constituent of the atmosphere, with any number of cosmogonically abundant and stable constituents such as Ar, Ne or N 2 as possible candidates for the background gas. Lewis (1971) had previously speculated on the possibility of a nitrogen atmosphere on Titan, result- R.H. Brown et al. (eds.), Titan from Cassini-Huygens, Springer Science +Business Media B.V

3 178 S.K. Atreya et al. ing from the photolysis of ammonia outgassed from the ices that were needed to explain Titan s low density. Danielson et al. (1973) analyzed the spectroscopic observations also, and concluded that the atmosphere was thin, comprising only a few millibars of CH 4. In the late 1970s, results of Titan s surface temperature began to emerge. Conklin et al. (1977) interpreted their 3 mm observations to imply a surface temperature of 200 K, which was surprisingly greater than Titan s black body radiation temperature of 82 K. This led Hunten (1978) to propose an atmosphere of 20 bar of N 2, whose pressure-induced opacity could explain the high surface temperature. In their pre-voyager model, Atreya et al. (1978) calculated that approximately 20 bar of N 2 could be produced initially on primordial Titan, from the photolysis of ammonia (NH 3 ) during the satellite s accretionary heating phase. Ground-based observations could not detect nitrogen, as N 2 has no rotational-vibrational spectral lines due to dipole transitions. Thus, by the end of the 1970s the situation with Titan s basic characteristics was less than satisfactory surface temperature ranged from 82 K to 200 K, surface pressure ranged from a few millibars to 20 bar, and the make up of the atmosphere from pure CH 4 to N 2 with traces of CH 4 and/or H 2 thrown in. This broad range defined the envelope of the engineering models used for guiding the Voyager 1 encounter with Titan. On 12 November 1980, Voyager 1 flew past Titan at a distance of 6490 km (3915 km from surface). For the first time, it detected nitrogen, found trace quantities of methane in the stratosphere, a thick haze blanketing the satellite which prevented composition and most other measurements below the lower stratosphere a pressure of 1496 ± 20 mb and a temperature of 94 ± 0.7 K at the surface, as well as numerous minor constituents in the stratosphere, including C 2 -hydrocarbons and nitriles (for a complete summary of the Voyager 1 and Voyager 2 findings, see Atreya 1986 ; Coustenis and Taylor 1999 ; Lorenz and Mitton 2008). N 2 was identified in its Rydberg, Lyman-Birge, and Lyman-Birge-Hopfield bands by the ultraviolet spectrometer (Broadfoot et al ; Strobel and Shemansky 1982) but placed only an upper limit on the argon abundance (see Section 7.2.1). The Voyager radio occultation experiment derived a mean molecular weight that was consistent with a pure nitrogen atmosphere, but the possibility of a small non-zero argon component could not be ruled out either (Eshleman 1982). This is because the radio occultation experiment derives the mean molecular weight, not composition. By itself, it yields only a relative pressure temperature profile from a determination of the atmospheric refractivity. The refractivity of the atmosphere as a function of time (height) during occultation is derived from the angle of refraction of the received frequency of the radio signal (S and X bands, respectively at GHz or 13 cm, and 8.6 GHz or 3.6 cm on Voyager). The relative pressure temperature profile can then be obtained from refractivity, which is related to the atmospheric composition (mean molecular weight). In order to obtain the absolute pressure temperature profile, an independent knowledge of either the temperature or the composition is required (Eshleman et al. 1977). The composition of the troposphere was not constrained at the time of the Voyager measurements. Therefore, the radio occultation experiments relied on the temperature measured by the Voyager infrared spectrometer at 1 mb in the stratosphere as a reference temperature. Any uncertainty in the reference value would propagate to the lower levels. This resulted in a slight uncertainty in the tropospheric and the surface temperatures of Titan. The resulting surface temperature had a range from 93 K to 95 K. The corresponding mean molecular weights were 27.8 and 28.3, respectively (Eshleman et al ; Lindal et al. 1983). The average of the two values led Eshleman (1982) to conclude that the mean molecular weight was consistent with a pure N 2 atmosphere. On the other hand, the higher value (28.3) could accommodate up to 4% Ar, whereas the smaller value (27.8) could be explained by including 2% CH 4. Since the composition of the troposphere was unknown, even the average value of the mean molecular weight (28) could accommodate within a small range various proportions of Ar (36 AMU) and CH 4 (16 AMU). Less than a year after the Voyager 1 flyby of Titan, Voyager 2 observed Titan on 26 August 1981, but from a much greater distance of 666,190 km. Despite its remoteness, Voyager 2 yielded valuable data on Titan s multiple haze layers and the photochemical species. The Voyager flybys gave a fleeting but tantalizing glimpse into the unusual and complex world of Titan. Their findings were instrumental in establishing the importance of studying Titan in detail, for it may even provide clues to the way Earth was before life emerged. The detailed composition of the gases and the aerosols, what lay beneath the thick haze, and the nature and composition of the surface were some of the mysteries of the time that could be revealed only by returning to Titan with an orbiter and an entry probe. A grass-roots effort by scientists in the United States and Europe in the mid 1980s culminated in what we know now as the Cassini Huygens mission, a joint undertaking of NASA, ESA and ASI (Italian Space Agency). The first ever in situ exploration of Titan was carried out by the ESA-furnished Huygens probe. After its release from the Cassini orbiter on 25 December 2004, the Huygens probe reached Titan on 14 January 2005, and began a historic 148-min descent through the moon s atmosphere and landed on the surface. Although not designed as a lander mission, the probe continued to send the composition data to the Cassini orbiter for another 72 min upon reaching the surface, and the carrier signal from the probe was recorded by ground-based radio telescopes for yet another 2 h and 2 min. This was a remarkable feat, considering the Huygens probe was expected to send data from the surface for only 3 min. Titan is also being explored remotely by the Cassini orbiter in

4 7 Volatile Origin and Cycles: Nitrogen and Methane the multiple close flybys of the moon since the arrival of the spacecraft at the Saturn System on 1 July As a consequence of these in situ and remote observations, new and vastly improved data have become available which form much of the basis of the discussions in this chapter. 7.2 Origin and Evolution of Titan s Nitrogen Atmosphere Two distinct possibilities existed for the origin of Titan s nitrogen before the Voyager observations: (1) N 2 is primordial, i.e. it was captured directly as N 2, or (2) N 2 is secondary, formed from the dissociation of ammonia, i.e. NH 3, not N 2, is primordial. The Voyager observations did not yield the data that could unambiguously answer the question of the origin of nitrogen on Titan. We had to wait another 25 years for Cassini Huygens to settle the issue. In the following paragraphs, we review the various possibilities and present the current status of the origin of nitrogen on Titan Direct Capture of N 2 Lewis and Prinn (1980) suggested that the dominant form of nitrogen in the solar nebula was N 2. Owen (1982) initially proposed that Titan s atmosphere resulted from the direct capture of the nebular material. In that case, its Ne/N 2 should be solar, or 6.2 (using N/H = for the solar nitrogen from Grevesse et al. 2005, but scaling the neon value up by 2.7 to account for the newer x-ray stellar data of Drake and Testa 2005, so that solar Ne/H = , as discussed in Table 1 of Atreya 2007b). However, neon has never been detected in Titan s atmosphere. The Voyager ultraviolet spectrometer (UVS) observations of Ne at 736 Å yielded a very low upper limit of 0.01 for Ne/N 2 above Titan s homopause (Strobel and Shemansky 1982). In the well-mixed homosphere, the Ne/N 2 would be even smaller, not greater, than this upper limit of This is because the concentration of species in the diffusive separation region above the homopause is controlled by their individual molecular weights (or, scale heights). This would result in greater mixing ratio of a lighter species, such as Ne (20 AMU, compared to the heavier background of ~28 AMU) above the homopause than in the well-mixed atmosphere below. Thus the Voyager UVS data imply that the Ne/N 2 mixing ratio on Titan is well below the solar value of 6.2. The same conclusion is reached by examining the tropospheric mean molecular weight, M. As determined by the Voyager radio occultation experiment, M is very close to 28 (Eshleman 1982), which also does not permit any significant quantities of neon to be 179 present in the atmosphere. Owen (1982) surmised that the way out of the neon dilemma is that N 2 was initially trapped in a hydrate, together with CH 4 and Ar. This idea was based on an earlier work of Miller (1961) who proposed that the ice in Saturn s moons is largely in the form of a methane hydrate (CH 4 6H 2 O), and of Lewis (1973) who suggested that the methane hydrate in the cooling solar nebula goes to completion at 60 K. Neon is highly volatile, and even at that temperature its vapor pressure is 40 bar, so that it could not have been trapped in the hydrate. Thus the non-detection of neon was not surprising, according to Owen (1982). If the above scenario of Titan s nitrogen atmosphere forming from the degassing of N 2 -CH 4 -Ar containing hydrate is correct, it would imply that solar proportions of argon and carbon must be present. The solar 36 Ar/N2 ratio is 0.11 (solar N/H = from Grevesse et al. (2005), but Ar/H = from Anders and Grevesse (1989), as the former used oxygen as proxy to derive Ar, and Ne whose correction is given in the previous paragraph). The Voyager UVS observations near a resonance feature of argon at 1048 Å yield an upper limit of 0.06 for 36 Ar/N 2 above the homopause (Strobel and Shemansky 1982). Unlike neon, the argon mole fraction could be larger in the homosphere. This is because 36 Ar is heavier than the background atmosphere of N 2, which would lead to a drop in the mole fraction of 36 Ar above the homopause due to molecular diffusive separation. Thus, the Voyager data on argon could neither support nor discard the direct capture of N 2 hypothesis of the origin of Titan s atmosphere. The interpretation based on CH 4 /N 2 was similarly ambiguous. The solar C/N is 4.07 (based on solar C/H and N/H from Grevesse et al. (2005)). The primordial carbon could be in the form of CO, CO 2, carbon grains or organic material. CO and CO 2 were found to be quite small in the atmosphere by Voyager. If most of the carbon were tied up in methane, then CH 4 /N 2 would be solar, or ~8. If indeed methane was so abundant on Titan, a huge reservoir of liquid methane would be present on the surface (the methane triple point of K is close to Titan s surface temperature of 94 K). This would lead to the collapse of the atmosphere, as nitrogen gas readily dissolves in liquid methane. The Voyager Infrared Interferometer Spectrometer Radiometer (IRIS) observations yielded CH 4 /N 2 = 0.02 at 100 mbar in the lower stratosphere (Hanel et al. 1981). Although no direct measurements of methane could be made below the tropopause, using the IRIS data and the Voyager radio refractivity data Samuelson et al. (1997) derived a 50% supersaturation of CH 4 in the upper troposphere and a CH 4 mole fraction close to 6% near the surface. Lacking the critical data on the distribution of methane in the troposphere and the mole fraction of primordial argon in the homosphere, the Voyager observations were not able to settle the question of the origin of nitrogen on Titan. There were still two competing hypotheses (a) the pre-voyager

5 180 S.K. Atreya et al. model of Atreya et al. (1978) and Hunten (1978) where nitrogen was acquired from the Saturnian subnebula in the form of NH 3, whose subsequent photolysis led to the formation of an N 2 atmosphere in Titan s warmer past, and (b) the post- Voyager model of Owen (1982) according to which nitrogen was captured directly. In a later publication, Owen (2000), however, agreed with the model of the formation of N 2 from NH 3 photolysis, considering that direct trapping of N 2 requires the temperature of planetesimals to be below 30 K, which is far too low for the Saturn subnebula where Titan formed N 2 as a Secondary Atmosphere from Primordial NH 3 Speculations of a nitrogen atmosphere resulting from the photolysis of ammonia were made as early as 1971 (Lewis 1971). Considering Titan s density of 1.8 g/cm 3, it is believed that Titan is made of roughly 40% ice by mass, and the rest rock. A small fraction of the ice is expected to have been in the form of ammonia ice initially. Prinn and Fegley (1981) argued that kinetic inhibition to the conversion of N 2 to NH 3 in the nebula and the subnebulae of the outer planets was minimal. Thus, NH 3, not N 2, was the dominant form of nitrogen in the Saturnian subnebula where Titan formed. Even before this reaffirmation, the idea of Titan s nitrogen as being a secondary atmosphere had caught on (Section 7.1). Three possibilities exist for producing N 2 from the dissociation of NH 3 : photolysis, impact, and endogenic N 2 from NH 3 Photolysis During the accretionary heating phase the volatiles were released from the ices to the atmosphere of Titan. Lunine and Stevenson (1987) suggested an atmosphere that contained copious quantities of methane, ammonia and water vapor. Atreya et al. (1978) developed a photochemical model, according to which the photodissociation of NH 3 takes place at wavelengths below 300 nm, and peaking at ~195 nm. Their photochemical scheme for ammonia is shown in Fig The photolysis of ammonia produces amidogen radicals (NH 2 ). About a third of ammonia so destroyed is recycled by the reaction of these radicals with hydrogen atoms. Nearly two-thirds of the NH 2 radicals recombine to form hydrazine (N H4 2 ) molecules (Fig. 7.1 ). Hydrazine acts essentially as the rate limiting step in the subsequent production of N 2. If the temperature is sufficiently warm, N 2 H 4 remains in the vapor phase and is dissociated by the solar ultraviolet flux to form hydrazyl radicals (N 2 H 3 ). The selfrecombination reaction of N 2 H 3 leads to the production of N 2 on Titan (Fig. 7.1 ). If the temperature is too low, little NH 3 Fig. 7.1 Photochemical production of N 2 from NH 3 in Titan's primordial atmosphere (updated from Atreya et al. 1978). Below 150 K, little NH 3 is in vapor phase, and whatever small intermediate product hydrazine (N 2 H 4 ) forms from it condenses, thus preventing N 2 formation. Above 250 K, water vapor has sufficiently large vapor pressure, so that OH from its photolysis reacts with NH 2 and NH 3 drastically decreasing the subsequent yield of N 2, despite some recycling of NH 3 by H from H 2 O would be present in the form of vapor, and whatever small amount of hydrazine is formed from it would condense and snow out of the atmosphere, preventing further chemistry and production of N 2 (impacts are ineffective in converting NH 3 or N 2 H 4 in any phase to N 2, as discussed in the next section). Atreya et al. (1978) found that the ideal temperature range is determined not just by the thermodynamic phase of N H4 2, but also the relative humidities of water and ammonia. At relatively high temperatures, a large quantity of water vapor coexists with the ammonia vapor. The photolysis of H 2 O vapor results in the highly reactive hydroxyl molecules (OH). The latter react with NH 2, severely limiting the formation of N H4 and the subsequent production of N 2. At low temperatures, hydrazine condenses and N 2 2 does not form, as discussed above. With these considerations, Atreya et al. (1978) determined that the ideal range of temperatures is K for the production of nitrogen from ammonia on Titan. Allowing for the increased solar ultraviolet flux in the pre-main sequence of the Sun (Zahnle and Walker 1982), Atreya et al. (1978) and Atreya (1986) calculated that up to 20 bar of N 2 could be produced in less than 2 Myr on primordial Titan initially, i.e. not accounting for subsequent N 2 escape. Atreya et al. (1978) assumed an isothermal primordial atmosphere in their nitrogen evolution model. For more realistic time scales for the formation of nitrogen, possible vertical gradients in the atmospheric temperature should be taken into account. Considering a range of possible abundances of H 2 O, CH 4 and NH 3 on primordial Titan (Lunine and Stevenson 1987), Adams (2006) developed a radiative transfer model for primordial Titan. Additional constraints on the pre-main sequence solar flux from x-ray to the infrared were included, based on the work of Ribas et al. (2005) and Zahnle and Walker (1982). The NH 3 photochemical scheme used was similar to the one used by Atreya et al. (1978). The use of the atmospheric temperature structure derived from radiative transfer led to longer time scales for

6 7 Volatile Origin and Cycles: Nitrogen and Methane the generation of nitrogen than the isothermal model of Atreya et al. (1978). The longer time scales are a consequence of the much smaller abundance of ammonia available in the photochemical regime of Titan s stratosphere because of cold trapping of the gas at the tropopause, which was not an issue for the isothermal models with temperatures far greater than the tropopause values. For example, current models require a little over 30 Myr to produce 10 bar of N 2 on primordial Titan. This will be discussed further in Section N 2 from Impacts High velocity impacts were suggested by Jones and Lewis (1987) and McKay et al. (1988) as a probable cause of Titan s atmosphere. Considering the possibility of shockinduced dissociation of ammonia from high velocity impacts during Titan s accretion, McKay et al. (1988) carried out a laboratory simulation experiment using a mm Nd-YAG laser on a mixture of NH 3 and CH 4. This process generated relatively large quantities of N 2. For a simplified accretion scenario, they could produce up to 25 bar on N 2 in the late stage of Titan s accretion. Although an attractive idea, the shock-induced production of N 2 encounters several difficulties. Nitrogen was not the only molecule produced in their experiment. The total yield of numerous hydrocarbons was even greater than N 2, i.e. up to 100 bar. H 2 production was nearly 4 bar. The H 2 amounts are so enormous that they exceed the present H 2 abundance a thousand-fold, unless the hydrogen escape rate throughout Titan s geologic history was also improbably large, exceeding 1000 times the present value (Waite et al. 2005) of cm 2 s 1 from the Ion and Neutral Mass Spectrometer (INMS). We are not aware of any mechanisms that could sustain such a rapid loss rate of H 2 from Titan. The large hydrocarbon production may also pose a further challenge, as most of them would end on the surface, depositing a layer 15 km thick. There is little direct evidence of such a thick aerosol layer on Titan s surface, although a substantial organic component to the surface material on Titan cannot be ruled out. Finally, the shock experiment did not include water vapor. As we discussed in the context of NH 3 photolysis, presence of water vapor would inhibit N 2 production (Section ), especially in the cooling phase of the shock. A variation on the impact-induced delivery of Titan s atmosphere was proposed by Griffith and Zahnle (1995). These authors suggested that cometary impacts resulted in Titan s nitrogen atmosphere. In this scenario, the source of nitrogen lies in the comets that condense out of the solar nebula, not in the planetesimals that condensed from the Saturnian subnebula to accrete Titan. In the cometary hypothesis, nitrogen is believed to have been supplied 181 originally in the form of complex organic molecules or molecular nitrogen. If comets were indeed the source of Titan s atmosphere, the D/H ratio on Titan should resemble that in comets. Although the D/H ratio has been determined in the ice of only a handful of comets (Halley Eberhardt et al ; Balsiger et al ; Hyakutake Bockelee-Morvan et al ; and Hale-Bopp Meier et al. 1998a), in each case its value was found to be larger than that in Titan s hydrogen reservoirs. The average value is found to be (3.16 ± 0.34) 10 4, which is greater than the D/H ratio either from CH 3 D (1.3 (+0.15, 0.11) 10 4 ; Bézard et al ; Coustenis et al. 2007) or H 2 ((2.3 ± 0.5) 10 4, Niemann et al. 2005) in Titan s troposphere. The cometary D/H has also been determined in one organic compound, HCN, and the value (2.3 ± 0.4) 10 3 (Meier et al. 1998b) is more than ten times greater than the above values on Titan. Thus, comets do not appear to be the main source of Titan s atmosphere. Finally, direct infusion of N 2 by comets is unlikely, as comets are deficient in N 2 since the temperature in the region of their condensation between the orbits of Uranus and Neptune was too high (50 60 K) for trapping nitrogen. We discuss the cometary hypothesis further in Section Endogenic N 2 Matson et al. (2007) have surmised that the dissociation of ammonia in Titan s interior may have produced some, if not all, of Titan s nitrogen, especially if catalyzed by clay or metals. If Titan formed 2.5 to 5.0 Myr after the calcium aluminum inclusions (CAIs) were created (as suggested for Iapetus, Castillo-Rogez et al. 2007) then differentiation occurred early and a stable core formed. During differentiation, heat from short-lived radioisotope decay and gravitational energy enabled rapid serpentinization of most of the silicate phase. (In serpentinization, water-rock reactions comprising oxidation and hydrolysis of low-silica and ultramafic silicate rocks such as olivine or pyroxene occurs, resulting in large quantities of heat, together with molecular hydrogen, serpentine, brucite and other minerals, as discussed in more detail in Section ) Water and chemical reactants were trapped by the hydrated silicate that accumulates into the core. After a few hundred million years, temperatures become high enough for ammonia decomposition to take place, producing molecular nitrogen and hydrogen. As Titan cooled, nitrogen was trapped in ice as a clathratehydrate. The outgassing of nitrogen occurred from time to time when the clathrates were destabilized. This scenario predicts that nitrogen is still being released from the interior. This would lead to a replenishment of nitrogen lost to space, and resetting of the 14 N/ 15 N isotope ratio to terrestrial or near-terrestrial value, as is the case with 12 C/ 13 C in the CH 4

7 182 S.K. Atreya et al. that is also believed to be released from Titan s interior to replenish the methane destroyed photochemically. The present data yield near-terrestrial 12 C/13 C, whereas 14 N/15N is severely depleted. At the moment, there is no simple solution to the isotope dilemma. In view of the above, the endogenic production of N 2 should be considered as tentative for the time being until further research is carried out. It is important to recognize also that any endogenic source would only supplement, not replace, the photolytic origin of nitrogen. This is because warmer Titan in the past would have certainly put large quantities of NH 3 vapor into the atmosphere, which could not have escaped photochemical destruction to produce nitrogen Origin and Evolution of Titan s Nitrogen Atmosphere: The Cassini Huygens Perspective The Cassini Huygens Mission collected the data that finally helped settle the question of the origin and evolution of the atmosphere on Titan. Two sets of measurements are particularly relevant: the primordial argon abundance, and the abundance of the nitrogen and carbon isotopes. None were available until they were done by Cassini Huygens in The abundance of primordial argon, 36 Ar, can shed light on the form of nitrogen acquired by Titan originally. The nitrogen isotope ratio is important for determining the mass of N 2 that must be formed originally to explain the present N 2 amount in the atmosphere. And, the comparison of the nitrogen isotope abundances along with the carbon isotope abundances provides clues to the evolutionary history of the gas. The Gas Chromatograph Mass Spectrometer (GCMS) on the Huygens probe measured the 14 N/ 15 N ratio in the troposphere of Titan by analyzing the N 2 data at 28 and 29 Da, i.e. 14 N 14 N and 15 N 14 N (Niemann et al. 2005). Similarly, the Ion and Neutral Mass Spectrometer on the Cassini orbiter measured the ratio at ~1000 km altitude in the thermosphere (Waite et al. 2005). Likewise, the 12 C/13 C was measured by both instruments from 12 CH 4 and 13 CH 4. Primordial argon, 36 Ar, as well as radiogenic argon, 40 Ar, produced from the radioactive decay of potassium, 40 K, in the rocks in the core, were also measured by the INMS and the GCMS. Table 7.1 lists these values. A comparison to the terrestrial values is given also, using the GCMS data. We use the GCMS data for the mixed atmosphere for this comparison instead of the INMS data, as the latter were measured in the thermosphere (~1000 km) where they are diffusively separated, so their extrapolation to well-mixed atmosphere is somewhat model-dependent. The mole fraction of 36 Ar is (Table 7.1 ), which is well below the solar 36 Ar/N 2 value of 0.11 (Section ), and 38 Ar was not even detected. This implies that nitrogen was not captured directly as N 2, nor did it arrive as N 2 from the subnebula, but as another nitrogen compound, most likely ammonia (NH 3 ). Had Titan acquired its nitrogen in the form of N 2, it would have also accreted large (i.e., solar) quantities of primordial argon, i.e. nearly 11% by volume, or over 300,000 times greater than measured, and over 1 million times greater when N 2 escape is included see below. During the accretionary heating phase, ammonia should have been present as vapor in the atmosphere of Titan. Subsequent photolysis of this ammonia then led to the nitrogen we find on Titan today, as discussed in Section The 14 N/ 15 N is found to be 183 in Titan s troposphere, which is ~0.67 times the value in the terrestrial atmosphere (Table 7.1 ). This implies loss of nitrogen from the atmosphere to space Table C/13C, 14N/15 N, D/H ratios, and the Ar, Ar, Kr, and Xe mole fractions ( f ) Ratio GCMSa in homosphere INMS b in heterosphere Titan/Earth (c) using GCMS 14N/15 N 183 ± 5 (from N 2 ) 240 ± 20 (1,450 km) ± 5 (1,000 km) 143 ± 5 (extrapolated to surface) 12C/ 13 C 82.3 ± ± 10 (1,450 km) ± 5 (1,000 km) 81.0 (extrapolated to surface) D/H (2.3 ± 0.5) f Ar (2.8 ± 0.3) 10 7 <(4 ± 1) f Ar (4.3 ± 0.1) 10 5 (1.26 ± 0.05) 10 5 (1050 km) f Kr, and f Xe <10 8 Not measured a GCMS measurements are from the Huygens probe from 146 km to the surface. Errors quoted represent 1 standard deviation (Niemann et al. 2005). The GCMS determined the D/H from HD and H 2. The value derived from the Cassini CIRS data using CH 3 D and CH 4 is D/H = (Bézard et al ; Coustenis et al. 2007). The 14 N/ 15 N from ground-based submillimeter observation of HCN is 60 ± 6 (Marten et al. 2002), confirmed by Cassini/CIRS HCN data (Vinatier et al. 2007). This is or about a factor of four less than the terrestrial value of 272, most likely due to the fractionation in photochemical production of HCN (Niemann et al. 2005). b INMS values are from Mandt et al. (2009), except for 40 Ar (Magee et al. 2009), revised from the values published by Waite et al. (2005) from the first flyby of the Cassini orbiter between 1174 km and 3000 km above Titan s surface. c The terrestrial inorganic 12 C/13 C = 89.45, is the Vienna Pee Dee Belemnite Standard.

8 7 Volatile Origin and Cycles: Nitrogen and Methane in Titan s past. Lammer et al. (2000) found that non-thermal escape due to solar wind ion pick-up and atmospheric sputtering could have resulted in a loss of up to 40 bar of N 2. However, this conclusion was based on a pre-cassini Huygens 14 N/ 15 N = 60 ± 6 derived from HCN (Marten et al. 2002). That nitrogen isotope ratio is a factor of three smaller than that in the principal reservoir of nitrogen, N 2 (Table 7.1), most likely due to fractionation in the process of photochemical formation of HCN. From a comparison of the current 14 N/ 15 N = 183 in N 2 in Titan s atmosphere measured by the Huygens GCMS with the terrestrial 14 N/ 15 N, Niemann et al. (2005) estimate that the mass of primitive atmosphere of Titan was between two and ten times today s value and perhaps several times the present mass was lost over geologic time, based on a model of diffusive separation of 15 N from 14 N by Lunine et al. (1999). Lammer et al. (2000) concur that diffusive separation is an important process, but ion pick-up and atmospheric sputtering caused by a high solar wind outflow during a Post T-Tauri phase of the Sun is required to explain the large nitrogen isotope anomaly. They estimate a loss of roughly 10 bar of atmosphere from Titan due to this process in the first 500 Myr, in general agreement with the earlier finding (Lunine et al. 1999). (Note that the mass loss estimates in both Lunine et al. and Lammer et al. were based on the then available smaller 14 N/ 15 N = 60 ± 6 from HCN, but the mass loss numbers given here have been scaled to reflect the larger GCMS value of 183.) Comparing the GCMS nitrogen isotope ratio with the terrestrial value implies that Titan would have started out with a nitrogen pressure of between 5 and 10 bar in its primordial atmosphere, in order to explain the present surface pressure of 1.5 bar. Considering a radiative transfer model for the primordial atmosphere of Titan, prevalent solar fluxes, and the photochemical scheme of Atreya et al. (1978) and Atreya (1986) for the production of N 2 from NH 3 (Section ), Adams (2006) calculated that it would take, respectively, 17 and 33 Myr to accumulate 5 and 10 bar of N 2 on primordial Titan. The time scale could be as low as Myr for 5 10 bar, if the methane abundance was 30% lower, because of the role of CH 4 in radiative transfer. As we discussed in Section , the ideal temperature range for photochemical conversion of NH 3 to N 2 is between 150 and 250 K. Post-accretionary evolution models show that Titan s temperature was in this range for Myr (Lunine and Stevenson 1985), which is more than adequate for producing the above initial inventory of nitrogen, even allowing for uncertainties in the primordial atmosphere models. If needed, the ammonia photolysis model is also capable of accommodating even much greater loss of nitrogen than the above 3 8 bar, considering the long duration of ideal temperatures for photolysis on primordial Titan. The nitrogen isotope ratio in Titan s atmosphere is also remarkable from another point of view. It seems to place Titan in the same family of solar system objects as the 183 terrestrial planets, in so far as the origin of nitrogen is concerned. Later we will see the similarities extend beyond nitrogen. In the right panel of Fig. 7.2, we illustrate this point by showing a comparison between Titan s 15 N/ 14 N isotope ratio with that of Venus, Mars and the Earth. The 15 N/ 14 N ratio in the atmosphere of Mars is larger than the solid Mars value as measured in the AL84001 meteorite, but is consistent with it after escape of nitrogen from the atmosphere is accounted for. Thus, the three terrestrial planets, Venus, Earth and (solid) Mars, have very similar 15 N/14 N ratios. The 15N/ 14 N ratio of Titan would also be similar to the terrestrial planet value if Titan lost some nitrogen as discussed above. The nitrogen in the atmospheres of Venus (3.5% by volume), Mars (2.7%) and the Earth (78%) is also secondary, i.e. it formed from the dissociation of a nitrogen molecule, most likely ammonia, in the early stages of planetary evolution. The Cassini Huygens observations show that similar to the terrestrial planets, Titan s nitrogen atmosphere is also secondary. Finally, a comparison of Titan s nitrogen isotope abundance with that on comets, solar wind and other primordial sources (left panel of Fig. 7.2 ) provides further constraints to the origin of nitrogen. Very few measurements exist for the comets, however. Furthermore, the nitrogen isotopes in more Fig. 7.2 Nitrogen isotope ratio in Titan s atmosphere compared with those on the terrestrial planets, and primary sources, including the interstellar medium (ISM), Sun (SW: solar wind, SOHO: Solar and Heliospheric Observatory), Jupiter (which represents the protosolar value) and comet Hale-Bopp. For Mars, two values are shown, in the atmosphere (ATM) and Allan Hills meteorite (ALH84001); the latter presumably represents the solid Mars value (prepared in consultation with Paul R. Mahaffy)

9 184 S.K. Atreya et al. than one molecule (two) nitrogen containing molecules have been determined for only one comet so far. Submillimeter observations of HCN in comet Hale-Bopp yield 14 N/ 15 N = 323 ± 46, or 15 N/ 14 N = for the central value (Jewitt et al. 1997). Another observation of the same comet gives a nearly identical value of 330 for 14 N/ 15 N (or, for 15N/ 14 N) from HCN (Ziurys et al. 1999), which is close to the terrestrial value of 14 N/ 15 N = 272 (or, for 15 N/ 14 N). On the other hand, optical data yield an 14 N/15 N = 140 ± 30 from CN in comets Hale-Bopp, LINEAR and a dozen other comets (Arpigny et al ; Hutsemékers et al. 2005). Waite et al. (2009 ) and Mandt et al. (2009) propose a cometary, rather than the subnebular (NH 3 ), origin for Titan s N 2, considering that the revised INMS value of 14 N/15 N =143 ± 5 in the homosphere of Titan matches the value of this ratio in the CN of a dozen comets. This is an interesting alternative, but several caveats should be noted. First, the INMS value of 14 N/ 15 N in the homosphere (143) is model-dependent, as it is derived by extrapolating the value measured in the heterosphere (above 900 km, where 14 N/ 15 N = 188), where diffusive separation dominates. Second, the derived INMS value for the homosphere (143) differs from the in situ determination of 14 N/ 15 N = 183 directly in the homosphere by the Huygens GCMS. The GCMS value does not resemble the cometary 14 N/ 15 N, being greater than that from CN and less than that from HCN. Third, the cometary 14 N/ 15 N ratio used for comparing the INMS derived 14 N/ 15 N for Titan s homosphere is not from the principal reservoir of nitrogen on comets, ammonia, or even the next most abundant one, HCN, but from a tertiary product, CN. The 14 N/ 15 N in HCN is more than twice that from CN in comet Hale- Bopp, the only comet for which the ratio has been determined in both HCN and CN ( 14 N/ 15 N from NH 3 is not yet available, but may be greater than that in HCN due to photofractionation). Fourth, it would take an unrealistically large number of impacts of comets to deliver the amount of nitrogen present on Titan. Finally, such impacts, or impacts of smaller number of larger comets delivering equivalent amount of N 2, would introduce excessive amounts of H 2 into Titan s atmosphere and result in a D/H value different from that observed on Titan, as discussed in Section The cometary nitrogen reservoir and its potential for delivering N 2 to Titan (third and fourth points above) are further elaborated in the following paragraphs. The dominant nitrogen bearing molecule on comets is ammonia, whose abundance is % relative to water (Bockelee-Morvan et al. 2004). The nitrogen isotope ratio in NH 3 is not measured. However, experience shows that 14N/ 15 N fractionation occurs in photochemistry, so that the ratio is smaller in the product than in the parent. This would imply that the 14 N/ 15 N in NH 3 of comets might be greater than that in HCN, a product of the chemistry between NH 3 and CH 4 of comets. 14 N/15 N = 323 from HCN in comet Hale- Bopp, the only comet in which the ratio has been determined. The cometary 14 N/ 15 N ratio from HCN is a factor of 2.3 greater than the INMS value for Titan s homosphere (143). On the other hand, 14 N/ 15 N = 140 ± 30 in CN of certain comets is similar to the extrapolated INMS value for the homosphere of Titan. Incidentally, photochemical fractionation of nitrogen isotopes is evident in comets as it is on Titan. The 14 N/ 15 N ratio in the cometary CN is a factor of 2.3 smaller than that in HCN; CN being a photoproduct of HCN (Hutsemékers et al ; Ziurys et al. 1999). Similarly, the 14 N/ 15 N ratio in Titan s main (parent) nitrogen reservoir, N 2 (143, using the revised INMS value), is also 2.3 times the value derived from HCN (60), a photochemical product of the N 2 -CH 4 chemistry. If comets were indeed the source of Titan s nitrogen, the nitrogen isotope ratio in its atmosphere should reflect that in the dominant nitrogen bearing molecules in the comet, NH 3 and HCN, not CN. The INMS 14 N/ 15 N ratio is a factor of 2.3 smaller than that in the cometary HCN, and probably even less compared to that in the cometary NH 3. If, for some mysterious reason, CN of comets were still responsible for producing Titan s N 2, the paragraph below gives an estimate of the magnitude of required cometary impacts. In comets, the abundance of HCN is 0.1% relative to H 2 O (Bockelee-Morvan et al. 2004). The abundance of CN has not been determined, but is expected to be much smaller than this, being a product of HCN. For example, Wilson and Atreya (2004) find CN/HCN = 10 3 from their photochemical model of Titan s atmosphere. In the most optimistic, but clearly unrealistic, scenario where all of the cometary HCN was converted to CN, the abundance of CN would be same as HCN, or 0.1% relative to water. A typical comet of 2 km diameter would then deliver kg of CN to Titan, taking the comet density to be same as that of water ice. The amount of N 2 in Titan s atmosphere today is kg, based on a 1500 mbar atmosphere containing 95% N 2 and 5% CH 4. Assuming that all of the cometary CN was converted to N 2 upon the comet s impact, again a stretch, it would take a staggering two billion 2 km diameter comets to deliver the amount of nitrogen currently present on Titan. In reality the number would be far greater. Since CN is expected to be only a small fraction of HCN, that not all of the CN is going to end up in N 2, and that the starting inventory of nitrogen was perhaps three to five times greater than that present on Titan today, the number of comets would easily exceed a trillion comets. Fewer, but bigger (than 2-km), comets delivering an equivalent amount of N 2 could be envisioned also. However, cometary impacts, whether in the form of large comets or smaller ones, whether recent or in the past, including those in the purported Late Heavy Bombardment between 3.8 and 4.1 Gyr ago, etc., would also deliver excessive H 2 and produce too large a D/H ratio, etc., as mentioned previously in this section and discussed in Section It should be noted also that the probability of a comet striking

10 7 Volatile Origin and Cycles: Nitrogen and Methane Titan is low. For example, the probability of a comet striking Mars, a somewhat larger object, is only once in 62 Myr. In summary, photolysis of ammonia on primordial Titan is found to be the most robust mechanism for producing Titan s nitrogen atmosphere. Nevertheless, further measurement of the chemical composition and isotope abundances of nitrogen and other elements in a multitude of comets together with modeling will benefit our understanding of the contribution of comets to the make-up of the planets and satellites in the solar system. 7.3 The Cycle of Methane on Titan The maintenance of a stable atmosphere of nitrogen on Titan is critically dependent on the warming resulting from the presence of its second most abundant atmospheric constituent, methane. In the absence of methane, Titan s nitrogen would condense and the atmosphere would collapse (Lorenz et al. 1997a, 1999). The photochemistry of methane produces hazes and molecular hydrogen, amongst other products. The hazes absorb the incoming solar infrared radiation and heat the stratosphere. The collision-induced opacity due to the H -N2, CH -N and N -N 2 collisions raises the temperature below. The resulting methane initiated warming of the atmosphere is crucial for preventing nitrogen from condensing out from the atmosphere as droplets. On the other hand, the stability of methane itself is not a given. Methane has a meteorological cycle in the troposphere that could lead to its loss to the interior, if only temporarily. In the stratosphere and the ionosphere, photochemistry destroys methane irreversibly. Episodic resupply of methane from its possible storage in the interior of Titan is necessary to restore it to levels 185 where it can participate in the atmospheric radiative balance and photochemistry, although the amount available to condense on to the surface in reservoirs such as the polar lakes may be highly variable. In this section, we will review this complex meteorological photochemical hydrogeochemical cycle of methane on Titan that is somewhat similar to the terrestrial hydrological cycle, and was referred to as the "methalogical cycle" by Atreya et al. (2006) The Meteorology of Methane For the first time, the vertical distribution of methane in Titan s troposphere was available following the in situ measurements made by the Huygens Gas Chromatograph Mass Spectrometer (Niemann et al. 2005). The GCMS determined the CH 4 mole fraction at a high spatial resolution starting at an altitude of 146 km altitude above the surface, and on the surface after the probe s landing. The distribution, shown in Fig. 7.3, is remarkable in many respects. First, the CH 4 mole fraction is uniform, with a value of 1.4 ± from the lower stratosphere (146 km) to ~32 km altitude which is just below the tropopause (40 km). Second, the mole fraction gradually increased in the troposphere until it reached a maximum value of 4.9 ± at 8 km, below which it remained constant down to a level just above the surface. Third, the methane signal suddenly increased as the probe descended through the 16 km altitude and remained high for a couple of km (see inset in Fig. 7.3 ). Finally, the methane mole fraction began to increase again after the probe landed on Titan, reaching a plateau within 2 min to a value that was ~50% greater than that just above the surface (not shown). The methane mole fraction then remained constant essen- Fig. 7.3 Methane mole fraction in Titan s atmosphere from the GCMS. The CH 4 mole fraction is 1.4 ± in the stratosphere. It increased below 32 km, reaching a plateau of 4.9 ± at about 8 km. The inset shows the behavior of CH 4 as the probe descended through km region. A sudden increase in the methane counting rate relative to N 2 (in this case 14 + N, which is used as a proxy for N 2 to get around the problem of saturation in the N 2 counts) was detected starting at 16 km and lasting over about 2 km. This was presumably caused by a droplet (or frozen droplet) entering the heated inlet of the GCMS as the probe descended through a cloudy/hazy region (Niemann et al. 2005). The haze apparently extended to at least 21 km where it was detected by the Huygens Descent Imager (DISR), but was first registered by the GCMS at ~16 km where its concentration was large enough for detection by the GCMS

11 186 S.K. Atreya et al. tially to the end of data transmission from the probe 1 h and 12 min later. The uniform mixing ratio of methane in the lower stratosphere is due to the fact that the unit optical depth, where photolysis destroys most of the methane, is much higher, i.e. between 700 and 800 km. The increase in the methane signal at 16 km indicates that the probe descended through a cloud or haze of methane that evaporated in the heated inlet of the GCMS. The increase of signal at 16 km does not imply either the beginning or end of the cloud/haze layer; it is just where the mass of CH 4 condensate most likely, a liquid or frozen droplet was large enough for it to be noticed by the GCMS. The increase in the CH 4 mole fraction after landing indicates moist soil from which methane was evaporated by the heated underbelly of the probe. A slight decrease in the methane mole fraction toward the end of the surface measurements was noticed also, indicating the pool of methane in the Huygens landing site was limited. The above behavior of methane is similar to the familiar behavior of water on the Earth. In both cases, convective ascent of a subsaturated air parcel from the surface to the lifting condensation level (LCL) results in saturation and condensation of the condensible species methane on Titan and water vapor on the Earth. Clouds form. Precipitation takes place. The surface becomes wet. Above the LCL, the mole fraction of the condensible species is limited to its saturation value, reaching the lowest point at the tropopause, and not rising above it in the warmer stratosphere. The relative humidity of methane in the Huygens landing site (10.3 S, W) was found to be ~50% just above the surface (temperature 93.8 K, CH 4 triple point K). The measured CH 4 mole fraction was constant to ~8 km altitude, as is expected in equilibrium. If the LCL were indeed at 8 km, the methane relative humidity should reach 100% at ~8 km. However, the relative humidity turns out to be only 80% considering CH 4 in equilibrium above pure liquid methane, as shown by solid line in Fig On the other hand, since nitrogen gas dissolves in liquid methane, the CH 4 saturation vapor pressure is lower over this binary mixture (Kouvaris and Flasar 1991 ; Thompson et al. 1992). As a consequence, methane in fact reaches 100% relative humidity at 8 km (dashed line in Fig. 7.4, Atreya et al. 2006), not 80%. This is where the condensation of methane is expected to begin. Above this altitude, methane would remain at 100% relative humidity, if equilibrium prevailed. This is just what is found up to ~15 km altitude (Fig. 7.4 ). The departure of the CH 4 data from the saturated CH 4 mole fraction above ~15 16 km is due to the transition of methane condensate from the liquid to the solid phase above this altitude, to which the formulation of Kouvaris and Flasar does not apply. The above discussion illustrates classic textbook example of equilibrium thermodynamics at work in Titan s atmosphere. The methane cloud/haze could actually extend well beyond the 16 km altitude where the GCMS first detected it. In fact, the Descent Imager (DISR) on the probe detected haze at ~21 km level (Tomasko et al. 2005), which is most likely the upper extent of the solid methane particles. Such a transition from liquid methane droplets below ~15 16 km to solid particles above is entirely consistent with the principles of equilibrium thermodynamics (Atreya et al 2006 ; Tokano et al. 2006). The detection of both an upper ice cloud of methane and a drizzle of liquid methane below has been reported by Adamkovics (2007) from their infrared obser- Fig. 7.4 GCMS measurements of the CH 4 mole fraction ( dots ) are shown against the saturation mole fraction curves for the CH 4 vapor over pure CH 4 liquid ( solid line ) and for the N 2 dissolved in the CH 4 liquid ( dashed line ; Kouvaris and Flasar 1991). The relative humidity of CH 4 is ~50% just above the surface. CH 4 is subsaturated up to 8 km, but reaches 100% relative humidity at this altitude, indicating the onset of condensation. The discrepancy between the data and model ( dashed curve ) above ~15 km is because of the change to solid phase of methane (after Atreya et al. 2006)

12 7 Volatile Origin and Cycles: Nitrogen and Methane vations in the equatorial region covering the continent of Xanadu. Based on the above cycle the average annual rainfall of methane in the equatorial region, such as the landing site of the Huygens probe, would amount to less than 10 cm per year, similar to the annual rainfall in terrestrial deserts. From time to time, convective storms could take place. Lorenz (2000) poses a global average energy limit on re-saturation of the atmosphere if a storm locally depletes it. The time interval between storms is estimated to be years, and could be much longer at a given location. Greater and more frequent precipitation is expected in the polar regions. Lakes have been inferred from the Cassini radar observations, largely at high northern latitudes of 78 N and higher (Stofan et al. 2007). The larger surface reservoir of liquid methane and the 2-4 K colder temperatures in the polar regions than the equatorial region (Flasar et al. 1981) result in ideal thermodynamic conditions for greater condensation and subsequent rainfall at the polar latitudes. The poleto-pole circulation of methane is expected to occur on time scales of years (Lunine and Atreya 2008) Clouds While it was speculated, largely on the near-coincidence of the Voyager-determined surface temperature with the triple point of methane, that methane might participate in a hydrological cycle as water does on Earth, there was no evidence to suggest an ongoing hydrometeorology on Titan until the late 1990s. Griffith et al. (1998) inferred the existence of clouds in Titan s upper troposphere from near-infrared ground-based spectroscopy acquired in This cloud system covered some 10% of Titan s disk, and may well be related to a cloud observed at nearly the same time at around 40 N by the Hubble Space Telescope (Lorenz and Mitton 2008 ; Lorenz 2008). No direct compositional determination of the cloud was made, but its altitude was consistent with that at which methane convective clouds might be expected to form. Further observations (Griffith et al. 2000) showed that the cloud cover varied on hourly timescales, suggesting that the clouds were actively convecting and/or precipitating. In Titan s thick atmosphere and low gravity, even the largest raindrops fall gently. Lorenz (1993) showed that drops controlled by the balance of surface tension and aerodynamic forces could grow to about 9.5 mm in diameter (compared with about 6.5 mm for water drops on Earth; this is a statistical limit, occasional giants can exist for short periods) but would fall at only 1.6 m s 1, compared with 9 m s 1 on Earth. Furthermore, the heat flux available to drive convection and evaporation from the surface in McKay et al. s (1989, 1991) radiative convective model is only ~1% of the incident sunlight, or about 0.04 Wm 2 on average 187 Fig. 7.5 Globally-averaged, annually-averaged energy balance of Titan s surface and atmosphere (original data from McKay et al. 1991). Fluxes are expressed as a percentage of the top-of-atmosphere insolation of 3.7 Wm 2. Solid arrows are shortwave (solar) fluxes; open arrows are the longwave (thermal) fluxes. Curved solid arrow indicates convective flux driving weather and clouds of ~0.04 Wm 2 (Fig. 7.5 ). Dividing by the latent heat of methane yields an average rainfall rate of only 1 cm per Earth year (the corresponding calculation for Earth yields, correctly, the typical rainfall of about 100 cm per year). These factors initially suggested (Lorenz and Lunine 1996) that rainfall would not be a major agent of erosion on Titan. However, Lorenz (2000) cautioned that, as with desert regions on Earth, landscapes can be fluvially dominated even with low average rates of precipitation, if that average rate is manifested by rare, but violent, rainstorms. Griffith et al. (1998) had noted that the Titan atmosphere holds several meters liquid equivalent of precipitable methane. Clouds began to be observed consistently on Titan from about 2001, as a result of several factors. Not least was the increased intensity of observation several large telescopes able to image Titan became available around this time, and the performance of their adaptive optics systems got progressively better. Compounding the telescopes availability was the establishment that Titan had time-variable weather, making it worthwhile to observe. Another factor was Titan itself, which cooperated by providing regular displays of large clouds around the south pole, which was at that time approaching midsummer. The association of the clouds with those latitudes where solar heating was strongest suggested that these clouds are at least driven by convection. Griffith et al. (2005) observed the cloud tops of evolving clouds at southern midlatitudes (41 61 S) in December 2005 (TB) using VIMS. Spectral data (the fit of the m m spectral region) indicate cloud-top heights and overall opacity. The evolution of the cloud-top heights suggest that small centers of vigorous updraft existed in the clouds, in

13 188 S.K. Atreya et al. some cases observed to be rising at 2 4 m s 1 in larger cloud systems, and 8 10 m s 1 in more compact (perhaps less wellmixed or younger) clouds. Spectra from Cassini s Visual and Infrared Mapping Spectrometer in 2005 revealed the presence of a vast tropospheric cloud on Titan at latitudes 51 to 68 N (Griffith et al. 2006) and the feature likely extended to the pole, although winter darkness prevented its observation. The derived characteristics indicate that this cloud is composed of ethane and forms as a result of stratospheric subsidence and the particularly cool conditions near the moon's north pole during the winter season. Preferential condensation of ethane, perhaps as ice, at Titan s poles during the winters may partially explain the lack of liquid ethane oceans on Titan s surface at middle and lower latitudes. Tokano et al. (2001) modeled the methane transport in the troposphere with a global circulation model including condensation. They suggest that the latitudinal distribution of methane humidity indicated by Voyager (2% mole fraction at high latitude, about 5% at low latitude, Samuelson et al. 1997) could not be reproduced by purely atmospheric processes, and that the surface must somehow influence the humidity. Cloud data were not given in this paper. A similar conclusion was reached by Rannou et al. (2006) who noted that the circulation would tend to desiccate low latitudes, yet the Huygens data had by this time confirmed the Voyager indication of ~5% methane mixing ratio at low latitude and thus a humidity of ~50%. Rannou et al. (2006) suggested that a low-latitude methane source was required. This model predicted cloud opacity as a function of latitude and season, notably suggesting persistent clouds at both poles in all seasons, and a seasonal midlatitude cloud in fall (e.g. at 40 N between 1985 and 1995). Ongoing Cassini and ground-based observations show that the south polar convective clouds seem to have ceased towards the end of Sporadic clouds have been seen since at a variety of latitudes these are discussed in the companion chapter by Lorenz et al. (2009) Methane and the Climate on Titan The climate of a planetary body reflects the balance between the absorption of sunlight, and the emission of thermal radiation to space. In the absence of atmospheric effects, these in turn depend on the body s distance from the Sun, and the optical reflectivity or albedo of the surface; the deviation from unity of the emissivity of most planetary surfaces is small and usually ignored. An atmosphere complicates the picture considerably, typically by reflecting and absorbing sunlight, and by blocking thermal emission. These factors are all significant at Titan, and were first quantified, analytically, by Samuelson (1983), who showed that the main roles of Titan s haze are to absorb blue light, while red and nearinfrared light is scattered and penetrates to the surface, and the gas in the dense atmosphere acts as a powerful greenhouse, absorbing thermal infrared radiation. A further analytic investigation of these effects for Titan and for the early Earth was made by McKay et al. (1999). The radiative balance was evaluated in much more detail with a numerical radiative convective model constructed by McKay et al. (1989) this considered 24 spectral intervals in the visible and near-infrared and some 46 intervals in the thermal infrared, including opacities for nitrogen, methane and hydrogen. The atmosphere was divided into 30 planeparallel layers, and a microphysics code calculated the optical properties of spherical tholin haze particles in each level. The effects of the haze profile, and methane absorption, allowed the model to be adjusted to fit the visible/near-ir spectrum of Titan measured from the ground, and the temperature profile was adjusted to achieve radiative balance. Convective adjustment was applied where the profile exceeded the local adiabat. The resultant profile could be compared with the temperature profile measured by the Voyager radiooccultation experiment. This model was used to show that condensation clouds of methane played a minimal role in Titan s overall radiative balance, a conclusion that has been borne out with time. This model, albeit a globally averaged one, allowed the competing effects (McKay et al. 1991) of antigreenhouse cooling by stratospheric haze (analogous to the hypothetical nuclear winter scenario for the Earth) and greenhouse warming in the troposphere to be evaluated in familiar terms. In essence, the effective temperature of Titan, which is determined by the distance from the Sun and the planetary albedo, is 82 K the antigreenhouse effect drops the temperature by 9 K, whereas the greenhouse increases it by 21 K, giving a net overall greenhouse warming of 12 K. Meanwhile it was recognized that several effects would drive Titan s climate with time. Not only did the solar luminosity change, forcing the climate directly, but also since methane is a condensible gas at Titan s conditions, a feedback exists wherein the amount of gas (and thus the strength of the greenhouse thermal opacity) depends on the temperature, and this feedback can magnify the effects of the solar forcing. Furthermore, over time, the relative amounts of methane and ethane in a surface-atmosphere reservoir would change as photolysis progressively converted methane to ethane. Additional radiative effects, depending on the photolysis rate, would include the amount of haze in the atmosphere, and the small but significant opacity due to molecular hydrogen, also produced by photolysis. An initial attempt to explore the solar forcing and volatile greenhouse effects was made by Lunine and Rizk (1989) who used a semianalytic grey radiative model. This effort

14 7 Volatile Origin and Cycles: Nitrogen and Methane found that temperatures on Titan several gigayears ago could have been rather lower, by K, than present, largely due to the photochemical evolution of the greenhouse inventory and slightly amplified by the solar luminosity variation. These effects were explored in more detail by McKay et al. (1993) who noted that the sensitivity of Titan s climate to solar luminosity and possible surface heating was considerably amplified by the existence of a large surface volatile reservoir. In other words, if Titan just had a few surface lakes, the surface temperature and pressure would change only modestly over the age of the solar system, while if there was a massive ocean, the pressure could have been much lower in the past (~0.2 bar 4 Gyr ago) than at the present (Fig. 7.6 ). Under these conditions, the ocean would have been frozen at about 74 K. McKay et al. (1993) kept presentday hydrogen greenhouse and haze production rates, and noted, but did not study, the further complication that under these cold early conditions, nitrogen ice clouds would form in the atmosphere and affect the radiative balance. The possibility that Titan s atmosphere may have been appreciably thinner in the past prompted consideration of possible geomorphological signatures of such history (e.g. Lorenz et al. 1995), in particular the increased abundance of small craters whose formation would be inhibited by atmospheric shielding at the present epoch (Engel et al. 1995). Further studies with the McKay et al. radiative convective model explored a couple of additional scenarios. The situation of Titan around a red-giant star (either our own, some 5 Gyr from now, or perhaps some Titan-like body around a Saturn-like planet around a red giant elsewhere in the galaxy) was explored Fig. 7.6 Possible history of Titan s atmospheric pressure over the age of the solar system, assuming photolytic conversion of a methane ocean and the production of haze and hydrogen at present rates (data from McKay et al. 1993). If there are only lakes of methane (with insufficient volume to dissolve much of the atmospheric inventory of N 2 ) the evolution is modest, but if a large ocean of methane existed, the pressure could have been much lower when the Sun was fainter several Gyr ago 189 in Lorenz et al. (1997a). Here, strong heating causes the atmosphere to puff up higher temperatures give a large scale height, and since haze is assumed to form at a fixed low pressure the haze column becomes very large and surface temperatures are not substantially enhanced. However, as the solar spectrum changes to admit more penetrating red light, and to produce less UV and thus produce less haze, surface temperatures can increase profoundly well above the ammonia-water melting point of 176 K even without invoking the enhanced greenhouse due to a large volatile reservoir. Lorenz et al. (1997b) also explored the climate in another direction with present-day and reduced solar luminosities, but with the methane abundance as an adjustable parameter to allow for the possibility that methane supply to the surfaceatmosphere system might be episodic, while its photochemical destruction is continuous. Thus there may have been episodes, perhaps many and sometimes long, when there was no methane available. This not only removes the CH 4 CH 4 and CH 4 N 2 collision-induced thermal opacity, but also (as hydrogen escapes over a subsequent period) would remove the N H2 2 greenhouse. These cooling effects are only partly offset by the loss of near-infrared CH 4 solar absorption and the clearing of the atmospheric haze: removal of methane drops the present-day surface temperature by about 5 K, although stratospheric temperatures fall dramatically from about 180 K today to around 60 K. Lorenz et al. (1999) made a set of convenient analytic fits to an ensemble of several hundred runs of the McKay et al. radiative-convective model, and reported the joint sensitivity of surface temperature to methane humidity and surface albedo: under present conditions, equilibrium surface temperatures rise by about 1 K for every 0.1 decrease in surface reflectivity, which might lead to some variation of surface temperatures across Titan s variegated surface. A drop of 0.1 in surface relative humidity from the present value of ~0.5 similarly results in a 1 K surface temperature drop. Also, temperatures drop by about 1 K for every 1 km increase in elevation. These analytic fits allowed further exploration of the evolution of the coupled surface-atmosphere climate system. Such a system (and the effect applies also to the Martian climate, where a substantial fraction of the total CO 2 inventory seasonally freezes onto the surface, as well as the water greenhouses on Earth and paleo-venus) will have an equilibrium where the surface temperature and the resultant atmosphere are self-consistent, in that the surface-atmosphere thermodynamic equilibrium P = f(t) and the radiative equilibrium T = f(p) curves cross, where T represents the surface temperature and P represents the partial pressures of the various gases. As first noted by McKay et al. (1993) a small inventory ( lakes ) is well behaved, having only one equilibrium (Fig. 7.7 ). However, if the volatile inventory of methane, ethane and nitrogen is large, there is the potential for strong positive feedback and the two curves form three equilibria of

15 190 S.K. Atreya et al. Fig. 7.7 Well-behaved surface-atmosphere equilibrium for a volatilepoor Titan (with ~0.1 bar of liquid at the present day, plus 1.5 bar atmosphere). The solid black curve with circles represents the surfaceatmosphere thermodynamic equilibrium, in effect the vapor pressure of the ocean as a function of temperature. The red dashed curve represents the radiative convective equilibrium, i.e. the temperature as a function of the greenhouse effect in the atmosphere for the present solar constant. Where the red and black curves cross ( marked with an arrow ) indicates a stable self-consistent equilibrium. The stippled blue curve shows the same for a solar constant of 80% the present value because the greenhouse curves are much shallower than the thermodynamic equilibrium curve, the crossing point moves only gradually as the solar constant increases (data from Lorenz et al. 1999) Fig. 7.8 Extreme climate sensitivity in a volatile-rich Titan (with 16 bar of volatiles). This figure is the same as figure 7.7, but with a larger volatile inventory. The corresponding vapor pressure curve ( black ) yields a thicker atmosphere for a given temperature, and the slope of the curve is very similar to the greenhouse curves. For a low solar constant ( blue ) there is a single stable self-consistent equilibrium. However, for the present solar constant ( red ) there are three equilibria. The central crossing is unstable, but two stable equilibria ( arrowed ) exist one at present conditions, the other with a much higher temperature and pressure. Small perturbations may cause the climate to flip between these conditions (data from Lorenz et al. 1999) which only the outer two are stable (Fig. 7.8 ). The choice of which of these two climates will Titan be in would depend on the history of the system, and indeed discontinuous jumps between them can occur a runaway greenhouse, or equivalently in the opposite direction, atmospheric collapse. This hysteresis makes climate prediction forward or backwards a significant challenge. Up to this point we have not considered variations of temperature with latitude. Such variations are important in the surface-atmosphere equilibrium, in that the planetwide strength of the greenhouse effect depends on the partial pressure of the various greenhouse gases, which will be close to thermodynamic equilibrium with the coldest part of the planet. Therefore, it is the coldest parts of the planet, and thus to a large extent the effectiveness of equator-pole heat transport that controls the planetary average temperature. In simple 1-D energy balance climate models used in the 1970s to study the stability of the terrestrial climate to lower insolation (so-called Budyko Sellers models) the heat transport embodying ocean as well as atmospheric motions is represented by a single heat transfer parameter D, which empirically today has a value of ~1 Wm 2 K 1. For Martian paleoclimate studies, the terrestrial value has been scaled by atmospheric pressure. For Titan, however, this approach yields a transport that is too high (Lorenz et al. 2001), and results in an equator-to-pole temperature gradient that is orders of magnitude lower than the 3 4 K suggested by Voyager observations. A much lower value, D~0.02 Wm 2 K 1 is required, and appears consistent with the idea that some climates, including the Earth's, may select combinations of heat transport modes in order to maximize the entropy production by the turbulent heat flux. While this idea is controversial, it is of interest that both the dimensions and the value itself of the heat transport parameter D are the same as the planetary entropy production ds/dt absorbed solar flux divided by absolute temperature. For the Earth, this is ~300 Wm 2 /300 K = 1 Wm 2 K 1, while for Titan the flux is ~3 Wm 2 and the temperature ~100 K, yielding ds/ dt~0.03 Wm 2 K 1. Even though its thermal inertia is such that the bulk of Titan s atmosphere does not warm and cool with the seasons, the surface temperature may see seasonal variations. These were first explored by Stevenson and Potter (1986) who suggested an analogy with the seasonal frost cycle on Mars. Under their scenario, some meters of methane might condense onto the cold winter pole (i.e. forming a transient liquid polar cap or seasonal lakes) before re-evaporating in spring. In reality, the seasonal variations (see chapter by Lorenz) are rather more complex, involving varying winds and humidity, as well as the availability of condensation nuclei. Nonetheless, the possibility of seasonal evaporation of surface methane lakes has been underscored by Mitri et al. (2007). Global circulation models of growing sophistication are being applied to understanding Titan s winds and, in particular, the seasonal and latitudinal variations in methane

16 7 Volatile Origin and Cycles: Nitrogen and Methane amount, clouds and precipitation. While these models are able to produce some basic features of the Titan climate such as dry conditions at low latitudes where dunes are seen (e.g. Mitchell 2008) and more abundant clouds at high latitudes (e.g. Rannou et al. 2006), many details remain to be resolved. Further, to date, there has been no latitudinally-resolved modeling of Titan paleoclimates with conditions different from the present Photochemical Destruction of Methane in the Stratosphere: The Ethane Ocean Dilemma The cycle of methane discussed in the above subsections represents at best a closed cycle of surface reservoirs, volatile evaporation, cloud formation, followed by precipitation, similar to the hydrological cycle on the Earth. In the worst case scenario, it may represent a net loss of methane due to surface runoff and loss to the interior of Titan through cracks and fissures in the surface. Any such loss is expected to be temporary, however, as various geological processes are likely to pump the methane back out to the surface over time. On the other hand, photochemistry destroys atmospheric methane irreversibly due to the escape of its product, hydrogen, as shown first by Strobel (1974). Consequently, the lifetime of methane in the atmosphere is ~30 Myr (Wilson and Atreya 2004; Yung et al. 1984). The UV photolysis occurs primarily above ~600 km in the stratosphere and is responsible for about one-third of the total CH 4 destruction rate of cm 2 s 1 ( cm 2 s 1 when referenced to the surface, E.H. Wilson, personal communication 2008), whereas the catalytic destruction by acetylene (C 2 H2 ) itself a product of the CH 4 photochemistry in the lower atmosphere accounts for the rest (Wilson and Atreya 2004). In each case, ethane (C 2 H 6 ) is the principal product, as seen in the chemical scheme below. (About 10 15% of the methane is destroyed in the ionosphere above 900 km, but little ethane is produced.) 2(CH 4 + h n CH 3 + H) and 2(C 2 H 2 + h n C 2 H + H) CH 3 + CH 3 + M C 2 H 6 + M 2(C 2 H + CH 4 C 2 H 2 + CH 3 ) CH 3 + CH 3 + M C 2 H 6 + M Net: 2CH4 C 2 H 6 + 2H Net: 2CH 4 C 2 H 6 + 2H [M is the concentration of the background atmosphere] Based on the above photochemical fate of methane, Lunine et al. (1983) predicted that over geologic time, a kilometer-deep ocean of ethane is expected to cover Titan s surface. This was a bold but reasonable prediction for its time, considering the limited state of knowledge about the 191 composition and structure of Titan s atmosphere. A quarter century later, observations made from the Cassini Huygens give no evidence of a global ethane ocean, however. The lack of ethane oceans may be attributed to a much more complex cycle of methane including photochemistry and geology than was previously envisioned and a better understanding of atmospheric vertical mixing. Current photochemical models show that many other products either compete with or are formed from ethane itself. Of special significance to the question of ethane oceans is the detection of benzene (C 6 H 6 ), which was not included in the model of Yung et al. (1984), which was the basis of Lunine et al. s prediction of global ethane oceans. Benzene was identified tentatively by the Infrared Space Observatory (ISO, Coustenis et al. 2003). Its presence at 1 4 ppbv levels in the stratosphere was confirmed by the Cassini Composite Infrared Spectrometer (CIRS, Coustenis et al. 2007), and at much higher mixing ratio of 1 5 ppmv above 900 km by the Ion and Neutral Mass Spectrometer (Waite et al. 2005, 2007). Despite these differences in the mixing ratio, the column abundance of C H6 6 in these two parts of the atmosphere is similar (Atreya 2007a), indicating potent production mechanisms throughout the atmosphere. Several pathways have been proposed for the formation of benzene in the neutral atmosphere and the upper ionosphere, but all can eventually be traced back to ethane. Benzene is the precursor to the formation of polycyclic aromatic hydrocarbons, PAH's, whose polymerization may result in some of Titan s most abundant haze layers. Both the ionosphere and the stratosphere contribute significantly to the haze. Thus, contrary to the earlier notions, ethane is not the end product (or even close to it) of the methane photochemistry, but it goes on to produce heavier hydrocarbons that subsequently form the multitude of haze and soot layers in Titan s stratosphere and higher (Wilson and Atreya 2004; Waite et al. 2007). Wilson and Atreya (2004) found that the inclusion of benzene and other chemical pathways in a coupled atmosphere-ionosphere photochemical model led to a reduction in the downward flux of ethane at the tropopause by a factor of 4 6 from a value of ~ cm 2 s 1 in the model that did not include them (Yung et al. 1984) and was the basis of earlier ethane ocean predictions. The new models thus predict much less condensation of ethane, reducing the depth of a global ethane ocean to m from the previous predictions of a kilometer. Atreya et al. (2006) found that the vertical distribution of ethane is also significant factor in determining the amount of ethane condensation. They found the best fit to the CIRS data on ethane requires substantially smaller eddy diffusion below 200 km than that in the previous models. The smaller stratospheric mixing seems consistent also with the conclusion of Yelle et al. (2008) who found a value of cm s for the eddy diffusion coefficient at the homopause, using

17 192 S.K. Atreya et al. 40 Ar as the observational constraint, or about an order of magnitude smaller than in pre-cassini models where the analysis was based on the CH 4 distribution alone (see chapter by Strobel et al. (2009 ) for a comprehensive discussion of composition and structure). Since the eddy diffusion coefficient varies inversely with some power of the atmospheric density, the lower values assumed by Atreya et al. (2006) for the lower atmosphere are reasonable. The reduced eddy mixing results in a further reduction in the downward condensation flux of ethane, by 30 50%. The reduction factor is probably larger, as the Atreya et al. (2006) model was done prior to the complete analysis of the upper atmospheric mixing from the INMS data (Yelle et al. 2008) that indicated a more severe reduction of eddy mixing in the upper atmosphere. Thus, with the above new constraints on atmospheric mixing and the composition from Cassini, the depth of a global ethane ocean is expected to be no greater than m, not a kilometer. Finally, geologic considerations indicate that the depth could be even smaller. Models of the interior show that methane on Titan may be outgassed only episodically, with the last such episode occurring 600 Myr ago (Tobie et al. 2006). This would further reduce the possible depth of the ethane ocean to less than 10 m over the geologic time. Since any ethane condensation would be gradual, it is quite likely that even a good fraction of this last bit may be largely sequestered as ethane clathrate in Titan s regolith (Atreya et al ; Lunine and Atreya 2008). Mousis and Schmitt (2008) estimate that a cryovolcanic icy crust of less than 2.3 km thick is required to bury all the ethane (and other less abundant hydrocarbon liquids) produced over the geologic time. Since the ethane rain ultimately originates in the stratosphere (Sections and 7.3.3), its sequestration as ethane-clathrate is expected to be at shallow depths just below the surface, unlike the methane clathrates that are predicted to be present in the high pressure ice some km below Titan s surface (Section 7.3.4). In conclusion, although roughly 40% of Titan s methane is expected to be converted to ethane over geologic time, global oceans of ethane are no longer predicted by current photochemical geological models, nor are they seen in the Cassini data. On smaller time scales, ethane may still condense in the atmosphere, especially at mid-high latitudes, and form lakes. Thermodynamic considerations predict a relative humidity of ethane ranging from 80% to 100% in the equatorial region, as shown in Fig. 7.9 (Atreya et al. 2008). Thus the condensation of ethane there would be at best marginal, and moreover any ethane drizzle may be sequestered in shallow subsurface clathrates. This seems to be the case in the Huygens landing site (10.3 S, W), where the probe mass spectrometer detected a slow evaporation of ethane from Titan s moist surface (Niemann et al. 2005). On the other hand, at polar latitudes, where the temperatures are Fig. 7.9 Photochemical profile of ethane with uncertainty envelope ( blue curve ) and its comparison with ( a ) the CIRS data at 5 S and 55 N ( vertical bars, with uncertainty envelopes ), and ( b ) C H saturation 2 6 vapor pressures, using the central HASI temperatures near the equator ( black curve ), HASI temperatures increased by 1 to accommodate the uncertainty in the temperature determination ( green curve ), and temperatures reduced by 2 relative to the HASI equatorial temperatures, to mimic high polar latitude conditions ( red curve ). The resulting C 2 H 6 relative humidity at the tropopause is % in the equatorial region, and over 200% at the polar latitudes (after Atreya et al. 2008) 2 4 cooler, ethane would almost certainly condense (Fig. 7.9 ). Since ethane liquid is fully miscible in the methane liquid, Titan s polar lakes are expected to be composed largely of a methane ethane mixture (Atreya et al ; Lunine and Atreya 2008). Brown et al. (2008) have, in fact, identified spectral features of ethane in Ontario Lacus (72 S, 183 W), which was observed by VIMS through Titan s atmospheric windows at 2.0, 2.7 and 5.0 m m during Cassini's 38th close flyby of the satellite. The authors conclude that ethane, probably in liquid solution with methane, nitrogen and other low-molecular mass hydrocarbons, is contained in Ontario Lacus. (The very large abundance of methane gas in the atmosphere prevents the identification of methane liquid on the surface.) Since the stratospheric mole fraction of ethane vapor is about a factor of 10 4 smaller than the CH 4 mole fraction, ethane liquid may constitute a small fraction of the methane ethane lakes at any given time. However, a realistic assessment of eventual fraction of the ethane liquid in the lakes is a complicated matter. Lorenz (1993) suggested that pure methane droplets may not survive their descent through Titan s troposphere. Even the survival of a ternary mixture of CH 4 -N 2 -C 2 H 6 droplets against evaporation depends on the relative humidity of C 2 H 6 (Graves et al. 2008). Until the distribution of ethane is determined from the lower stratosphere down to the surface, the fractional amount of ethane in the polar lakes will continue to remain highly uncertain.